DRAFT PLEASE DO NOT QUOTE

Transcription

1 Locally embedded FDI? Relations between sub-national labour market and economic development institutions across four advanced economies Phil Almond and Maria Gonzalez Menendez (presenting authors), Anthony Ferner, Patrick Gunnigle, Jonathan Lavelle, Sinead Monaghan, Gregor Murray and Olga Tregaskis contact: Paper prepared for the 16 th World Congress of the International Labour and Employment Relations Association (ILERA), Philadelphia, PA., 2 nd -5 th July, 2012 DRAFT PLEASE DO NOT QUOTE Affiliations: Almond, Ferner De Montfort University, UK Gonzalez Menendez University of Oviedo, Spain Gunnigle, Lavelle, Monaghan, University of Limerick, Ireland Murray, University of Montreal, Canada Tregaskis, University of East Anglia, UK 2012 Phil Almond, and co-authors

2 Locally embedded FDI? Relations between sub-national labour market, economic development institutions and foreign direct investors across four advanced economies 1) Introduction This paper is based on interim results for an ongoing, cross-national research project into the links between foreign-owned multinational firms, economic governance actors present at a sub-national level in their host countries, and issues relating to the governance of labour markets. The current paper covers regions in four advanced economies (Canada, Ireland, Spain and the UK) 1. Overall, the project seeks to explore three sets of questions: - First, it seeks to identify the key state and civil society governance actors and how they engage with multinational corporations (MNCs). Typical types of actor investigated include local/regional economic development and political bodies, the training and education sector, social partners, and so on. - Second, it looks at whether and how the supply and demand of skills and competencies by foreign MNCs is shaped by the above relationships, and what governance actors do in different national and sub-national contexts in order to improve the supply of skills, whether these are firm-specific or more generic. - Third, it asks whether variations in these relationships, both within and between national political economies, have the potential to affect the extent to which MNC units can be said to be locally or regionally embedded. In particular, it asks how differences in the governance actor-mnc relationship affect the ability of firms to maintain and improve positions in value chains, and to participate in the endogenous creation of skills that might drive wider economic development within the region or locality. Conceptually, this paper uses the model of a Sub-National Business System (SNBS), which adapts previous frameworks developed within the comparative capitalisms literature (e.g. Whitley (1999), for an application to MNCs see Kristensen and Morgan (2007) to subnational levels (for a fuller exposition, see Almond (2011a). Empirically, we draw on evidence from qualitative interviews of governance actors and of MNC subsidiary managers and directors, in two regions for each of the four national economies mentioned above. These countries were selected for their broadly comparable positions within the international competition for foreign direct investment (FDI). Each has become relatively highly dependent on FDI over the course of the last three decades. Equally, they were previously able to compete for FDI with their geographically closest trading partners the USA, in the case of Canada, and the European Union (EU) in the other three national economies at least 1 The regions covered in the current paper, and the relevant local researchers are: East Midlands and North West of England (Phil Almond, Anthony Ferner and Olga Tregaskis); Asturias (Maria Gonzalez Menendez and David Luque Balbona University of Oviedo) and the Madrid region (Javier Quintanilla, IESE) in Spain; the Shannon and Gaeltacht regions of Ireland (Patrick Gunnigle, Jonathan Lavelle, Sinead Monaghan); and the Kitchener-Waterloo sub-region (Tod Rutherford, Syracuse) and Québec province of Canada (Gregor Murray, Matthieu Pelard, University of Montreal). 2

3 partially on the basis of relatively low labour costs 2, light labour regulation and/or low corporate taxation. However, due to various changes in the global political economy of FDI briefly highlighted below each has come under pressure to offer more positive advantages, in the areas of infrastructure, innovation and skills ecosystems, which require a greater degree of positive institutional construction. Our aim is therefore to explore how these pressures have translated, through the prism of national and sub-national differences in the organisation of what has been called the state-mnc nexus (Phelps and Fuller 2001), into attempts to capture and retain investment. As internationally comparative work at this level is scarce, the current paper s primary concern is with identifying relevant actors, what sub-national actors do, the extent to which this varies according to the socio-political construction of sub-national business systems, and therefore the extent to which sub-national business systems represent a significant level of MNC-governance actor relations, and with what potential consequences for the analysis of industrial relations and human resource management in foreign-based subsidiary units of MNCs. Below, we present in slightly more detail the rationale and conceptual framework for the research. After a discussion of methodology, we then present early findings. As the research had not, at the time of writing, been completed in all of the four countries, these should be taken as highly provisional. Finally, the concluding section assesses some of the broader implications of this research for discussions on MNCs and employment systems. 2) Literature and conceptual resources As is well known, the literature on industrial relations and human resource management within MNCs frequently discusses host country effects (e.g. Ferner et al 2001; Bae et al 2001; Tuselmann et al 2010). Very broadly, this type of research deals with two concepts, albeit often within the same research projects. Firstly, and at least until very recently predominantly, this literature has examined the existence of imperative regulation in host countries particularly in the areas of collective employment relations and individual labour legislation which might be seen as constraints on foreign multinationals acting within the host country. Such constraints render difficult, or sometimes impossible, the transfer of country-of-origin - or global best practice inspired HRM policies. Such literature typically looks at whether and how MNCs, or particular types of nationalities of MNCs, conform with, seek to avoid, or seek to negotiate, the nature of host country imperative constraints. Second, particularly where based on a more serious engagement with the comparative capitalisms literature (Whitley 1999), some literature examines how MNCs may attempt to use institutionally embedded resources. These may include those derived from skills institutions, research and development infrastructures, or the presence of clusters of competent firms to engage as suppliers (e.g. Kristensen and Morgan 2007). Such resources are increasingly seen in this literature as the domain in which older, higher cost industrialised economies compete for FDI, through what has been termed institutionally-derived societal comparative advantage (Sorge 1991). The host effects literature has overwhelmingly dealt with national-level effects. Yet both constraints and resources can also vary at sub-national levels. With regard to constraints, 2 In the case of Canada, the production cost advantage with the Northern US states has typically derived more from exchange rates and the socialisation of health costs, rather than direct wage costs (Rutherford and Holmes 2011) 3

4 there is, in some cases, formal sub-national variation in the nature of business and employment regulation as applied to MNCs. Examples include the Right To Work states in the USA, or Special Economic Zones in developing countries and some Eastern European states, while countries such as Canada and Australia have some degree of internal variability in industrial relations frameworks. Meanwhile, Locke (1992) has shown the importance of regional political economy in affecting HR environments in different parts of Italy, and Elger and Smith s work (2005) on Japanese MNC units shows clearly that their peripheral employment relationships in the UK were facilitated by low unionization and lack of alternative employment in the relevant local labour market. Similarly, greenfield inward investors who initially opted for higher-road variants of non-unionism in the UK often located in areas remote from the traditional patterns of British employment relations (Hudson, 1994). Regarding resources, clearly nation states do not normally constitute homogenous economic spaces. Existing economic structures within localities and regions shape the labour markets from which firms, including MNCs, can draw. There are also possibilities for governance actors within regions more actively to attempt to shape the resources available to inward investors, through localised productivity coalitions, strategic attempts to attract specific types of public and private investment, and through enabling desirable forms of coordination between the various business system actors present in their geographical space. In summary, if institutional arrangements contribute to the competitive advantage of MNEs (e.g. Bélanger and Trudeau 2007; Kristensen and Zeitlin 2005), then MNEs may seek advantages derived from sub-national arrangements as well as those from the national level (Meyer et al 2011; Phelps 2000). Where there are advantages to localised supply chains, or collaboration with local firms or research institutions, the importance of supportive sub-national institutions is further increased (Porter 1998; Maskell and Malmberg 1999; Gertler 2003). As the subsidiaries of most MNCs outside obvious market-driven investment in retail etc. substantially recruit from sub-national labour markets, developing combinations of human capabilities which are difficult to replicate is likely to be a priority for sub-national governance actors in attempting to embed the operations of MNCs in their area. Equally, it is obvious that the effects of MNCs investment decisions, and the impact they have on employment and organisational practices in host economies, are mostly keenly felt at local and regional levels. Thus regionally important MNCs are likely to have substantial interest in the nature of sub-national business and employment systems, and may also have significant power to affect governance at local levels. Local educational institutions (Beugelsdijk and Cornet 2002), as well as financial incentives, aftercare and other resources commonly transmitted through sub-national agencies and/or regional governments (Fuller and Phelps 2004) are likely to be important here. Sub-national variation in elements of business systems is clearly not new. However, a range of literature suggests that it may, or should, have increased in recent years, and gained in importance. This includes: a) Macro-level work in political economy, particularly that influenced by the Régulation School (e.g. Jessop 1993), discussing the adaptations made by states in response to neoliberal globalisation and the crisis of Fordism. This broadly suggests that the broad shift away from state protection of national firms towards the promotion of competitiveness what Jessop (1993) summarises as the Schumpeterian state and the consequent concentration on securing viable positions in the global competition for mobile productive investment has, in our terms, led to host-country effects being less and less about imperative regulation/constraints, and more and more about the competitive provision of resources, a process which is said to be easier to enact at more local levels than the national. This is said to lead to new forms of multi-level governance (Brenner 1999), with important elements of 4

5 state authority transferred to supranational institutions (the World Trade Organisation, the EU), and elements of responsibility for institutional provision transferred to local/regional levels of state competence. b) Meso-level work in economic geography (e.g. Maskell and Malmberg 1999), which broadly argues that globalisation processes have led to tacit knowledge, said to be embedded in localised networks, being key to the economic competitiveness of MNCs and their subsidiaries. c) Work on the increasing fragmentation of global production networks (e.g. Rutherford and Holmes 2008, Herrigel and Zeitlin 2011), which implies an increasing sophistication in how lead firms make location decisions, with obvious effects on supply chain firms. This again would require a more localised concern with the competitive provision of resources than was hitherto the case. These arguments are pursued in more detail in an already published conceptual paper (Almond 2011a). Table 1, below, provides a brief summary of these arguments. 5

6 Table 1 Summary of change tendencies affecting sub-national/mnc relations Change tendency Causes Modified by Effects Spatial nature of economic governance: movement of authority to supra-national levels, and of responsibilities to subnational levels Crisis of Fordism Attempts to secure supply of innovation within broadly neoliberal globalization Pre-existing nature of sub-national vs. national political and economic governance Spatial fix setting nature of business systems moves upwards (to supra-national) and downwards (to sub-national) actors Provision of some host resources moves down to subnational levels, with increased potential for powerful MNCs and their local units to affect nature of host systems Knowledge economy Increased importance of (tacit) knowledge to international competitiveness Belief that tacit knowledge is easier to transfer on a local basis Capacity of MNCs to transmit tacit knowledge across geographically dispersed organizations Continuing importance of cost and market reasons for MNC location Attempts to create relatively localized ecosystems within which geographical spaces can compete on value and innovation Increased importance of localized knowledge resources may promote increased cooperation within host regions, i.e. increased embeddedness of host units in sub-national geographies Spatial and organizational fragmentation of production Technology, globalization and shareholder value considerations Continuing importance of large MNC corporate hierarchies Move away from branch plants to need to imbricate MNC units (and other local businesses) in global production networks Increased need for flexible high skills ecosystems, involving MNCs, governance actors and other local firms in production networks Source: Almond (2011a: 537) 3) Methodology The research is based on an internationally comparative cross-sectional research design across two regions in each of Ireland, the UK 3, Spain and Canada. Two regions were investigated in each national economy in order to permit an evaluation of the potential for intra-national as well as international difference. 3 It should be noted that, more precisely, our fieldwork was conducted in two regions of England. Economic and political governance of the types investigated in this paper is largely devolved to the parliaments/assemblies of Scotland, Wales and Northern Ireland, thus from this point we will deliberately refer to England except where discussing macro-level features of the political economy of the UK as a whole. 6

7 The first challenge was the identification of regions, whose boundaries can be somewhat fluid. Attempting to match regions in terms of size was quickly rejected, given the different scales of population in the countries involved for example many of the recognisable regions of England have populations in excess of the entire national population of Ireland - and different structures of sub-national political and economic governance. We therefore concentrated on the top level of sub-national governance in place. For Spain and England, this did not produce major difficulties. In the case of Spain, which has a clear structure of regional political government, two of the 17 autonomous communities were chosen (Asturias, an old industrial and mining region in a peripheral geographical location on the North Coast, and the Madrid region). For the UK, we selected two top-level regions within England, on which the then-labour government had instituted regional development agencies (RDAs), responsible for a range of meso-level governance activities, including the attraction and retention of FDI. We deliberately chose one Northern region with a relatively high level of regional identity (for England), and one Midlands region where regional rather than local identity is very weak. In the case of Ireland, there were somewhat more difficulties. Statistical regions, essentially a creation in order to comply with EU regulations, had little reflection on the ground. We therefore chose to investigate two specific regions where there was a history of meso-level economic development institutions Shannon, where much of the early inward FDI to Ireland was concentrated, and the rural Gaeltacht region. Finally, in Canada, the highest sub-federal level, and from secondary evidence and early fieldwork the most relevant actor locations in the state-mnc nexus, is clearly that of the province. This presents certain methodological locations due to the size of the provinces, both in terms of population but also in their vast geographical scope. In the event, something of a compromise was reached here: the most interesting intra-national comparison was clearly at a provincial level, and particularly between Quebec known to have a somewhat less free-market approach to the political economy that the rest of Canada (Bélanger and Trudeau 2007) and Ontario. With this, however, in order to capture networks based on proximity, the research concentrates on specific regions of provinces (note that in Canada, the hierarchy is federal state province region municipality). This, in turn, is easier in Quebec, which, partly because of its uncertain degree of affiliation to the Canadian federation, has pursued its own economic development strategy dividing the nation-province into seventeen regions. In Ontario, the idea of sub-provincial regions has not been pursued by the provincial government, and de facto regions such as the Kitchener-Waterloo-Cambridge Technology Triangle in which much of our research to date has been based have emerged on a more organic basis. Each country research team generated a profile of the sub-national business system within each location, generating a database of key state and civil institutional actors who had a functional presence within the sub-national location and an evident affiliation to current or prospective FDI within the sub-national location. Initially, all institutions were specific FDIrelated institutions as identified by primary and secondary sources of official national and sub-national documentation. These generally related to formal institutions, such as legal, economic, political and administrative systems (Mudambi and Navarra, 2002). Pilot interviews were held with key institutional actors within the location, predominantly regional development agencies who are considered the main interlocutor to MNCs in a location to ensure all relevant institutions were included. Following the interviews, actors were asked to identify other institutional actors within the sub-national location who engage with FDI. The result of this process ensured that all active sub-national institutional actors were identified.. Additionally, unanticipated institutional change during the course of the English research led 7

8 us to interview a number of individuals with new or changed roles towards the end of the project. Prior to the data collection stage of this research, a standardised semi-structured interview format was collectively designed by all research teams, pertaining to the interaction of institutional actors with FDI within the sub-national location. Two broad interview formats were devised to account for the different perspectives of the sub-national institutional actors and the MNC respondents. Specific dimensions of the interview schedule included role within the region, direct and proximate interactions with sub-national institutional actors, HR and labour markets and perceived strategically important MNCs within the location. The interview format also provided sufficient flexibility and scope to explore unique country- or sub-national-specific issues which may arise during the course of the interview. Unless the participant objected, interviews were recorded for accuracy and transparency. No individual, organisation or institution was identified, and anonymity was guaranteed. A standardised coding structure for qualitative interviews was also verified by all research teams prior to data collection. As this research is cross-national and inter-regional, the data analysis required a more robust and coordinated approach. In terms of analysis, a set of themes were deduced from primary literature to provide an initial framework to guide analysis. However, all research teams were encouraged to identify unique themes within their findings. Firstly, qualitative data from each sub-national location was to be collated, coded and analysed. Following this, each country was to compare the data from both sub-national locations to identify national level similarities and variance. Finally, data from all six sub-national locations were to be considered in their totality. This process is at different stages in the different regions: The English and Irish research is complete, while, for reasons of funding horizons, the Spanish and Canadian research is still ongoing. Thus, in terms of the latter stages of the analysis, this paper and presentation represent early, provisional efforts. In all, 151 interviews had been conducted by February in England, 57 in Ireland, 31 in Spain and 12 in Canada. The preliminary findings below also draw on secondary data, both that which is publicly available, and material given to us by respondents. Findings This section is divided into two main sections. First, the extended Table 2, below, will summarise findings to date on the institutional complexes involved and their relations with MNCs. Second, we will make some provisional comments on the nature of MNC-subnational governance relations, and on the idea of embedding foreign MNC units. 8

9 Table 2: sub-national governance and MNCs in regions of England, Ireland, Spain and Canada England Ireland Spain Canada Broad variety of capitalism Liberal market economy limited attempts at the creation of employer-led (but effectively state-sponsored) coordination ( ) largely abandoned by current government with very strong liberal orientation Liberal market economy, albeit with recent experience of nationallevel concertation. Productive economy dominated by largely US FDI Described as Mediterranean variety of capitalism, formerly state/family business dominated. Fairly dense industrial relations system, and regulated (though rapidly liberalising ) labour market, but in reality MNCs possess substantial ability to negotiate outcomes of regulation at local level. Broadly a liberal market economy by any European standards, but noticeably less liberal in the context of North America., particularly in the case of Quebec. Locus of competitiveness in FDI attraction Historically, linguistic and political links to USA. Relatively large domestic market. Importance of London-based financial services sector. Open economy. 1980s, relatively low labour costs and light labour regulation within European single market, attracting Asian tiger FDI. From late 1990s, light labour regulation retained by neo-liberal social democratic government, but increased concentration on securing the supply of innovation. Historically, linguistic and cultural links to USA, and cost competitiveness within European Union. Liberalised trade policies, low corporation tax, strength of education system, activism of national inward investment agency Formerly reliant on low labour costs within European market, rather than light regulation/low tax. Position undermined by EU expansion and trade liberalisation. Well-educated (and underemployed) labour force. General production cost advantage compared to Northern USA (exchange rate and low health insurance benefit costs) Superior quality of general education and vocational training Federal immigration policies important in attracting highly educated and skilled workers (esp. to Ontario) Retains knowledge economy advantages in certain sectors, postcrisis strategy somewhat unclear Degree of FDI focus Very high from 1980s, perhaps somewhat moderated since Extremely high, note political consensus on low corporation tax, post crash and bailout, including on Republican left. Since 1980s, in practice varies somewhat by region. In our regions, consistently high in Asturias, more moderate - in terms of public effort Perhaps somewhat more moderate than in most European cases. Concern to develop Quebecois 9

10 Highest ratio of employment in foreign affiliates in both manufacturing and services in OECD (OECD 2010; UNCTAD 2007) - in Madrid firms in Quebec. Degree of sub-national autonomy Highest sub-national level: Very low - regional policy entirely run by national government : Regional Development Agencies (RDAs, see below), important in regional coordination but with minimal regional democratically accountability (governance without government), except in London. Sub-regional 2010: Creation of Local Enterprise Partnerships (LEPs), effectively run by collectives of local authorities. Extremely limited funding Very low strongly centralised administration. Regional and local government fragmented, and operate with very limited autonomy or power Regional bodies have next to no governance role, other than coordinating EU funding. Regional development associations in two regions with particular circumstances (Shannon and Gaeltacht) supplement work of national development agency, but are currently subordinate to it. Highest sub-national level Elected regional governments have very substantial devolved powers within overall national framework, including economic development (but also education, transport, social assistance, etc.) Regional government interpretation of common labour law (e.g. permission for redundancies has been easier to obtain in right-wing Madrid than in left-wing Asturias) Local government of some relevance through ownership of industrial parks, role in regulating land usage etc, but lacks resources. Highest sub-national levels Extremely high level of autonomy at top-level provincial level over most domestic affairs; provincial governments are the key actors with regard to FDI governance policy Lower levels Sub-provincial regions have very little autonomy politically. Collectives of local authorities (and/or creations of provincial government s own, centrallydriven, regional policy in the case of Quebec). Local authorities important economic development and inward investment actors, especially in large Northern cities Regional politics and development orientations Regional politics per se of no importance due to centralisation at national level Some evidence of greater tradition of local state-civil society-firm tradition in Northern than Midlands region, possibly related to stronger regional identity, or history of (relative) local authority developmentalism. Regional policy has resulted in a proliferation of institutions at regional/local level yet these bodies have limited autonomy and decision-making power Regional government more active economic co-ordinator in Asturias than Madrid (business-friendly social democrat vs. liberal approach) note Madrid regional government an outlier in Spain in terms of its economic liberalism. Asturias pattern shaped by strong regional identity (although not nationalist), geographically selfcontained area The main locus of politics in Quebec is separatism/federalism. This is important for two reasons: a) Attempts to reinforce a Quebec national identity extend to the field of meso-economic coordination b) Electorally, conventional conservative parties blocked from power Quebec consequently has pursued 10

11 Regional social pact, formalising role of regional development agency, but also, e.g., knowledge economy/society strategy, employment/training. History of subsidy-chasing investment (1980s/90s), attempts to rectify this: a number of governance actors refer to attempts to align investment with regional priorities and cluster-building initiatives, e.g. attempts, with some success, to attract ICT services MNCs. (at least in a North American context) a relatively socialdemocratic interpretation of macro neo-liberalism. Higher institutional density of industrial relations system in Quebec than in Ontario. Ontario pursues a more conventionally liberal approach; fewer attempts at any systematic coordination Role of inward investment and other development agencies Inward investment: Attraction of MNCs divided between RDAs and UKnational body UKTI, with some role also for local-authority funded sub-regional agencies. Investor development largely in hands of RDAs, except in cases of largest investors which had (and retain) relations with national civil servants. RDA investor development officials significant soft governance actors in coordinating skills provision on adhoc basis, drawing up bids for skills-related funding, etc : Responsibility for attraction passes to private sector contractor working for UK-wide UKTI. No regional-level role, or regional focus in UKTI work. Local/sub-regional authorities remain free to engage in FDIrelated activity, often through LEPs, but budgetary climate highly unfavourable. Investor development Strong role for national inward investment agency at international, sectoral and local level Key role for indigenous development agency in a) Mediating the support and success of FDI to indigenous companies in terms of skills transfer and/or investment b) Facilitating a strong subsupply network of companies c) Encouraging partnerships with MNCs to leverage funding for research d) Supporting the activity of the national inward investment agency with specialist skills in human resource development and technology Ambiguous role of regional 11 Asturias: Regional development agency IDEPA seeks to attract FDI, work with existing MNCs, and improve competitiveness of regional industries by improving design, innovation, quality... nominally tripartite, but on a day to day basis substantially autonomous from social partners political appointment: personnel and approach can change for party political reasons. Madrid inward investment agency somewhat less active (and newer). Auto sector in Madrid mentions much less institutional support than in other regions, and reactive (by threat) rather than proactive (aftercare) Clustering initiatives (Asturias) within remit of inward investment agency Inward investment Inward investment agencies have a relatively restrained role in both provinces; remit does not extend beyond the attraction of investment. Sub-provincial offices of provincial body exist in Quebec, but with no real autonomy. Local investment agencies (e.g. Canada s Technology Triangle ) emerge on an ad-hoc basis in Ontario. Broader regional economic coordination Institutionalised tripartite relations on issues related to skills, organised on a sectoral level (Quebec), more ad-hoc and sporadic initiatives in Ontario. Existence of Quebecois regional policy leads to institutionalised support for local inter-firm cooperation in localised sectors in the form of a cluster-type strategy. Montreal has its own cluster

12 work of RDAs largely unreplaced. Broader regional economic coordination RDAs had strategic responsibilities in wide range of areas (innovation, skills, sustainability, transport) Also funded regional productivity alliances (particularly in Northern region). development agencies given the success of national development agencies. Coordination of skills etc largely takes place through a national agency. organisation independently of this. Importance of community colleges (further education) in ensuring competitiveness of manual workforces in Ontario. Sub-regional LEPs have inherited some of these roles, on an ad-hoc basis, but national orientation away from social-democratic style coordination Other governance actors Unions excluded. Employers organisations also absent. Employer-led (but, in practice, largely state-funded) institutions such as (national) Sector Skills Councils intended to aggregate skills demands and inform state skills strategy. Have some regional presence, but coordinating capacity weak due to general lack of employer involvement. Other skills institutions very unstable. Relatively autonomous and business-friendly higher education sector important in skills supply. Strong role for educational institutions in both the availability and supply of labour and also in research and development initiatives Unions have a limited role and tend to have little interaction with MNCs or other governance actors For strategic issues, e.g. matters that threaten investment/are to do with major new investment, the interlocutor is the regional civil service directly, rather than development agencies. These are political appointments, meaning social democratic in Asturias vs. much more free market in Madrid. Social partners involvement in social pacts etc (not Madrid), but no real involvement in inward investment policy MNCs are quite notorious in Asturias (regardless of actual weight in jobs) and get 4th Estate pressure, perhaps result of them being symbols of regeneration? Unions and employers organisations important in official skills system in Quebec, but not in cluster-type initiatives (provincial state-employer coordination), or inward investment. Quebec unions also control large investment funds, used to support economic development. Densely-institutionalised skills system in Quebec, less so in Ontario (but skill levels still historically very high, especially compared to USA). Autonomous and businessfriendly higher education sector important in skills supply and research collaboration. Evidence on MNCs and systemic Generally little interaction. Relatively limited at regional level, although some MNCs have taken 12 Some degree of involvement, primarily from formerly Some evidence of MNC involvement in clusters, and in

13 coordination Units in active, predictable and cyclical competition for investment well networked among relevant skills and political actors. Managers here relatively supportive of RDAs. Other large units saw utility of RDAs in simplifying complex institutional systems, esp. with regard to skills. prominent roles in national employment system at times. nationalised firms in foreign ownership. Generally little involvement with official skills system as to apprenticeships (other than for graduates) because of cost pressures formal skills system in Quebec (early stages of research) Some evidence of local activism, largely on individual managers initiatives/local CSR Evidence on MNCs and firm-level coordination Extremely little evidence of firm to firm dialogue outside supply chain. Sectoral productivity alliances (cofunded by RDAs) gave some degree of coordination. Evidence of inter-firm networking based on personal relationships and experience of the FDI sector Strong sub-supply sector with active links to FDI Asturias Active embeddedness of MNCs more through associational governance (e.g. their role in the Quality Club, in attempting to improve regional supply chains, HR networking among MNCs which are not directly related Clearer evidence than elsewhere, in some firms, of local managers acting strategically for the interests of the regional operation. See above Less evidence so far for Madrid. Additional sources: Almond et al (2011), Monaghan (forthcoming), Rutherford and Holmes (2011) 13

14 The Spanish and Canadian cases somewhat approximate to ideas of substantive sub-national business systems within an overall national pattern which is more defined in Spain than it is between Quebec and Ontario mainly because of the strength of regional/provincial government. This is driven by the political dynamics of regional politics and the threat of separatism in both nations. For those seeking to attract and embed FDI, this allows a degree of local flexibility, which is probably particularly important in the relatively densely regulated national business system of Spain. This allows regional actors to create institutionalised patterns of cooperation which augment the national business system, but at the same time sometimes risks creating zero-sum competition between regions. This is probably less of a risk in Canada due to its geographic scale. Ireland, on the other hand, retains many of the FDI-facing advantages of a liberal market economy, while, through a strong and strategic consensus on the importance of attracting and retaining investment, overcoming many problems of coordination through relatively informal networking both at national and more local levels. The capacity of this network is closely related to an interpretation of Irish (economic) national identity that is strongly FDI-oriented; the very weight of FDI within the Irish national economy creates mutually-reinforcing effects in this regard. Finally, in the much larger UK economy, the picture is more mixed. While there is national consensus on the importance of FDI to the economy, there is a lack of political consensus on how either direct delivery of services to MNEs specifically, or regional economic governance in general, should be organised. While the UK remains an attractive destination for FDI both because of the importance of London as a financial/corporate centre, because of a relatively large domestic market, and possibly still because of light regulation, coordinating actors in England often lack a strong geographic locus of activity and identity, whether within relatively autonomous sub-national business systems (as in Spain), or primarily through the orientation of the national business system (as in Ireland). Provisional comments on embeddedness Clearly, research which seeks to investigate the sub-national embeddedness of MNC units has to take account of a wider range of factors than the nature of governance systems. Drawing on early fieldwork in which the lead author participated, in all four countries, the framework in Figure 1 was developed.

15 Figure 1: Summary of factors affecting sub-national host embeddedness Skills and knowledge MNC need to access geographically-specific combinations of skills and knowledge. Value chain Need and possibility for subsidiary unit to bid for new investment in higher positions within global value chains Parent orientation Ability of MNC to engage with high trust relations with local firm and governance actors (shaped by dominant conventions of trust within MNC) Host national business system Difficulties of coordination in compartmentalised business systems Difficulties of ensuring adaptability in systems with greater degree of non-market coordination MODIFIED BY Degree of sub-national business system autonomy Higher levels of autonomy (political, but also of other business system actors) may create greater scope for institutional innovation DEGREE OF IMPORTANCE OF HOST SUB-NATIONAL BUSINESS SYSTEM TO MNC UNIT source: Almond (2011a) These include factors related to the MNC - the nature and specificity of skills and knowledge (Foss and Pedersen 2000), the nature of the value chain (Phelps and Fuller 2001), parent orientation (as possibly shaped by country of origin business systems, cf. Morgan and Kristensen 2006) as well as those related to the host national business system and degree (and nature) of sub-national business system autonomy. Our MNC interviews cover a range of companies which, in any given set of sub-national institutions, would be expected to have different levels of material interest in establishing active links, given their differences in terms of skills needs, value chain nature and degree of real mobility. Skills and knowledge Some of the economic geography seems to assume that the need to access locally-specfic knowledge is a given 4. In fact, as should be obvious, multinationals 4 This is perhaps the result of an exaggerated trust in resource-based views of the firm (and of the region), which assume the intra-organisational, and intra-network transfer of knowledge, is the raison d être of MNCs, rather than merely a potential source of competitive advantage. 15

16 vary considerably in their need to access geographically-specific combinations of skills and knowledge. It is important to examine the specific labour market of the particular MNC unit this may be primarily local/regional, but it may be inter-regional 5 (which creates potential problems of coordination, not least in the area of skill provision and funding), national, or even in some cases international. For most industrial MNC units, the labour market, exceptsometimes for senior managers, is primarily local. If a manufacturing unit requires particular skills in its manual workforce, it is therefore relatively reliance on local/regional labour market institutions, and host sub-national resources potentially gain in importance relative to host country effects (or, more simply, a firm with one production site in a country of any size is more concerned with what is available on the local labour market to overall national host country resources). This can mean that, in large industrial MNCs with, for example, plants in a variety of European countries, the different regions in which plants already exist compete for new investment through skills funding. This is very typically the case in the auto industry, for example. So, for example, one company we visited which is a subsidiary of a large French defence company is located in a small town in the North West of England because aerospace skills are strong there due to the historic importance of the formerly nationalised British Aerospace. This is in spite of the fact that its main customer is located in a different region. In this sort of case, there is strong potential for regional institutions to embed competitive advantage in global production networks through institution building in the skills area. As one Spanish trade union official told us, in attempting to explain Asturias s relative success in attracting FDI, compared to most other peripheral Spanish regions, an industrial working class is not created overnight. Here, regional skills alliances, and/or sub-national-state-sponsored clustering-type initiatives such as those developed in Quebec, can be of importance. But it is important to remember that this is not always the case. Where skills are generic, regional institutions are less important. Also, the resource-based view of the MNC and its regional interactions typically ignores the fact that some very high skilled MNC units draw on a much wider labour market and are less reliant on the local labour market. For example, the pharmaceutical firm Astra Zeneca had until recently an R&D facility in the East Midlands, but had very few links with local or regional institutions as it drew upon a national and international labour market for scientists, and its style of knowledge management meant that there was little overspill to the wider region 6. The value chain the crucial question here is to what extent does the subsidiary unit need to bid for new investment, and can it improve its value chain position. Here, we saw an interesting contrast in the English research between two large auto assembly plants, one owned by an American auto firm, the other by a Japanese firm. The UK managers of the Japanese plant know that, at least in the medium term, the parent firm s relative lack of production capacity within the EU market means that all new models in a certain market segment will be manufactured at their plant, and they do not have to bid for new investment. In the American firm, however, each new model is the subject of a very competitive bidding 5 This was particularly an issue in England, where despite the large size of regions in terms of population, long-distance commuting is common. The hollowing-out of British manufacturing (and the relative lack of natural boundaries between regions in the industrial heartlands of England mean that cross-regional commuting for the relatively well-paid jobs offered by large manufacturing MNCs is not at all unusual. At the other extreme, Asturias, bordered by mountains and the sea, experiences virtually no cross-border commuting of this kind. This may be non-trivial, in the sense that more contained labour markets are, all other things being equal, possibly easier to coordinate. 6 In fact, one director of a regional business incubation facility speculated that the closure of this plant had probably increased overspill to the region, by causing the development of new, smaller businesses in the life sciences sector. 16

17 process between half a dozen European plants, perhaps all of which are under threat of closure should the decision go the wrong way. So although the Japanese-owned plant is, at least in the medium-term, more secure, the managers of the American plant engage much more intensively with regional and national skills actors, politicians and development agencies to ensure their competitive position. They also have made strong efforts to localise their supply chain, going somewhat against the centralising initiatives of the US HQ, through, for example, encouraging Chinese suppliers to the global firm to locate in their region, in order to gain competitive advantage versus their European subsidiary competitors. These examples show that the frequently-used term embeddedness has to be treated carefully. Active engagement with local and regional bodies is not the same thing as security of longterm investment, and indeed, as in the US auto case, engagement may arise because of a lack of security, rather than the other way around. Also, many firms are, at least for the moment, firmly anchored in particular spaces in order to service markets, without necessarily having any great connections to regional governance actors. More broadly, any analysis of FDI and embeddedness/relations needs to ask, on a unit by unit basis, why the firm is in a particular location, whether it is doing something there that is unique and/or is local resource dependent, whether there are sister plants elsewhere in the world, and what the real degree of potential mobility of the production/service provision of particular units is. Parent orientation This project did not set out to examine country of origin effects (see Almond 2011b). However, it remains the case that all MNCs are not the same in terms of their ability to engage in high-trust, long-term relations with local firms and with governance actors, and that this is frequently related to the national business system of the parent company (Ferner 1997). While not a core basis of our comparison here, we have examples from both England and from Spain which might suggest that relative shareholder patience, and/or capacity to work within institutionally dense systems (in the case of some, though not all, German firms in Spain), sometimes remains a factor in both the capacity for subsidiary unit managers to deal autonomously with local institutions, and occasionally for parentcountry managers to seek to use local institutional systems, particularly of industrial relations, to their own ends of social partnership etc. Host national business system Obviously, a focus on the sub-national should not mean neglect of the national host system, which, as seen, at the very least sets the framework within which the sub-national business system works, and in the cases of England and Ireland goes much further than this. As is well known, scholars of comparative capitalism in advanced economies often differentiate between countries in which non-market coordination of capitalism is commonplace, such as most of continental Europe, or, in somewhat different ways, Japan or Korea, and more liberal, or compartmentalised business systems, such as most of the English speaking world. From our English results, we can confirm that getting companies to coordinate with each other on matters such as skills is extraordinarily difficult. While this is hardly an original observation, the extent to which this was true remained surprising to the English research team. For example, one regional development agency director encountered a situation where four fairly large chemical firms in the same industrial park had the same skills deficits, which were potentially threatening future investment, but where unit managers had not talked to each other about this. The regional development agency managed to coordinate a training programme, but this example illustrates how state intervention, which firms in liberal economies strongly distrust in general, is necessary for skills coordination in such economies. Specifically talking about the UK, employers associations are very weak 17

18 because they are not required for collective bargaining purposes, and this can undermine skills initiatives. On a preliminary basis, skills do not seem to be anything like as much of an issue for MNC managers in Ireland: whether this is to do with better national coordination (Ireland has a history of national social actor collaboration), to the education and training system itself, or to more exogenous factors (such as a relatively young workforce and the high level of managerial competence and networking in a relatively patriotic MNC sector, due to the critical importance of MNCs in the Irish economy), is a matter for further consideration. Coordination on future skills needs seemed to be relatively developed in Québec, which in spite of being in the fairly liberal national economy of Canada has a high number of active business and employment institutions at provincial level. Finally, Southern Ontario, while not particularly coordinated in a conventional sense, does seem to have developed means of ensuring micro-coordination between particular firms and vocational training institutions, certainly in comparison to the US Great Lakes regions with which it competes (also see Rutherford 2011). Again, research here is at an early stage, and this requires further investigation. Finally, we argue that high levels of sub-national business system autonomy may create greater scope for institutional innovation. Looking at our four countries, in England regional autonomy is very weak. The previous Labour government created regional development agencies, which were important both in dealing directly with MNCs, and in wider economic governance, but these had very weak political legitimacy, and did not map onto labour market areas well. The current government has abolished these, meaning there is now, in England, no serious political or economic governance between the local authority and national state levels. This, we believe, will create serious difficulties of coordination in England, and is likely to exaggerate already strong economic differentials between the London travel to work area and the rest of the country. Ireland also has weak regional governance, but a strong national economic identity of managers, and a relatively high capacity for private networking, means private governance solutions can often be found on an ad-hoc basis. Spain has very highly autonomous, democratically-elected regional institutions, and in some regions, such as Asturias, regional social pacts between employers, the region state and trade unions. This has the advantage that it is easier to mobilise local social and business actors around regional economic issues, but the disadvantage that inter-regional competition is so strong that multinationals are able to regime-shop between regions. Finally, Canada s federal system, when pushed to its limits in the Quebec province, does seem to enable effective interactions at provincial level, although one problem may be the fact that the provinces of Canada are so large that sub-provincial regional policies are necessary to deal with MNC units skills needs as related to their specific labour market areas. Concluding discussion There is obvious room for debate about the possibilities and sustainability of relations between sub-national governance actors and MNCs. As Rutherford and Holmes (2011: 3) indicate, while some (e.g. Kristensen and Zeitlin 2005) stress the synergies between TNCs and regional institutions, others view this relationship as much more tenuous and contradictory and ultimately subject to power asymmetries favouring (MNCs). Equally, as host effects move from imperative regulation to the competitive provision of resources for potentially mobile investment, we see that, for such mobile firms, formal stage regulation of labour markets is increasingly being over-ridden. This is not a simple case of deregulation, but rather of the supplementing, and sometimes challenging, of formal regulation with a complex, opaque network of governance actors, which may be public, para-public or private. 18